数据提取方法:单篇手稿内部报告差异分析及实用建议
该文聚焦系统评价与系统评价概览的数据提取环节,指出同一手稿内部也可能存在无法同时为真的报告差异。作者描述了单篇来源中内部报告差异的实例,旨在提高系统评价作者对此类方法学问题的认识,并建议由两名评价者独立提取数据以增加发现差异的机会,同时提供了处理不同类型不一致及报告此类不一致的建议。
要点
- 系统评价与系统评价概览的作者应核查同一研究内多处报告的信息是否一致并进行比较。
- 由两名评价者独立提取数据可增加发现差异的机会(若存在差异)。
- 作者提供了处理不同类型不一致及在系统评价中报告此类不一致的建议。
核心结论
- 系统评价与系统评价概览的作者应核查同一研究内多处报告的信息是否一致并进行比较。
- 由两名评价者独立提取数据可增加发现差异的机会(若存在差异)。
- 作者提供了处理不同类型不一致及在开展系统评价与系统评价概览时报告此类不一致的建议。
研究设计与样本
- 本文为方法学分析,描述单篇来源中可发现的内部报告差异实例,未报告样本量。
主要发现
- 报告差异被定义为无法同时为真的成对陈述。
- 系统评价与系统评价概览的作者在提取数据时可能遇到多来源间的报告差异,包括手稿与会议摘要之间、手稿与临床试验注册库之间。
- 此类差异也可在发表于科学期刊的单篇手稿内部发现。
临床怎么用
- 作者建议系统评价与系统评价概览的作者核查同一研究内多处报告的信息是否一致并进行比较。
- 作者建议由两名评价者独立提取数据以增加发现差异的机会(若存在差异)。
- 作者提供了处理不同类型不一致及报告此类不一致的建议。
局限性
- 原文未报告局限性
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: Data extraction from reports about experimental or observational studies is a crucial methodological step informing evidence syntheses, such as systematic reviews (SRs) and overviews of SRs. Reporting discrepancies were defined as pairs of statements that could not both be true. Authors of SRs and overviews of SRs can encounter reporting discrepancies among multiple sources when extracting data-a manuscript and a conference abstract, and a manuscript and a clinical trial registry. However, these discrepancies can also be found within a single manuscript published in a scientific journal. OBJECTIVES: Hereby, we describe examples of internal reporting discrepancies that can be found in a single source, with the aim of raising awareness among authors of SRs and overviews of SRs about such potential methodological issues. CONCLUSIONS: Authors of SRs and overviews of SRs should check whether the same information is reported in multiple places within a study and compare that information. Independent data extraction by two reviewers increases the chance of finding discrepancies, if they exist. We provide advice on how to deal with different types of discordances and how to report such discordances when conducting SRs and overviews of SRs.
引用
- Puljak L, Riva N, Parmelli E, et al. Data extraction methods: an analysis of internal reporting discrepancies in single manuscripts and practical advice. J Clin Epidemiol. 2019. PMID: 31541692. doi:10.1016/j.jclinepi.2019.09.003
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。